JobMesh

Senior Research Scientist (Multimodal Large Language Model) - PICO

ByteDance · San Jose, California, US

About the Team PICO-MR team is dedicated to pioneering core technologies for intelligent human-computer interaction in MR environments, with a focus on integ...

Job description

About the Team PICO-MR team is dedicated to pioneering core technologies for intelligent human-computer interaction in MR environments, with a focus on integrating multimodal large language models (MLLM) and tool-use capabilities to redefine user experiences. Our R&D directions cover cutting-edge fields including multimodal scene understanding, MLLM-based agent systems, tool-augmented MR interaction, 3D environment perception, and AIGC-driven content generation. Within MR scenarios, our work spans: MLLM optimization and adaptation for MR, intelligent task execution with tool use, multimodal scene understanding (vision, point clouds, text), AIGC-based scene generation, depth estimation (Mono/Stereo/MVS), 3D environment perception, large-scale 3D scene reconstruction (3DGS, NeRF, etc.), visual localization, and lighting estimation—encompassing both fundamental research breakthroughs and industrial-grade solution deployment. Responsibilities: 1. Lead the R&D of multimodal large language models (MLLM) tailored for MR scenarios, integrating vision, point clouds, text, and other multimodal information—including model architecture optimization, cross-modal alignment, data construction, ev...